[Truncated abstract] During the last fifty years mathematical models of catchment hydrology have been widely developed and used for hydrologic forecasting, design and water resources management. Most of these models need large numbers of parameters to represent the flow generation process. The model parameters are estimated through calibration techniques and often lead to ‘unrealistic’ values due to structural error in the model formulations. This thesis presents a new strategy for developing catchment hydrology models for representing streamflow and salinity generation processes. The strategy seeks to ‘learn from data’ in order to specify a conceptual framework that is appropriate for the particular space and time scale under consideration. Initially, the conceptual framework is developed by considering large space and time scales. The space and time scales are then progressively reduced and conceptual model complexity systematically increased until ultimately, an adequate simulation of daily streamflow and salinity is achieved. This strategy leads to identification of a few key physically meaningful parameters, most of which can be estimated a priori and with minimal or no calibration. Initially, the annual streamflow data from ten experimental catchments (control and cleared for agriculture) were analysed. The streamflow increased in two phases: (i) immediately after clearing due to reduced evapotranspiration, and (ii) through an increase in stream zone saturated area. The annual evapotranspiration losses from native vegetation and pasture, the ‘excess’ water (resulting from reduced transpiration after land use change), runoff and deep storage were estimated by a simple water balance model. The model parameters are obtained a priori without calibration. The annual model was then elaborated by analysing the monthly rainfall-runoff, groundwater and soil moisture data from four experimental catchments. Ernies (control, fully forested) and Lemon (53% cleared) catchments are located in zone with a mean annual rainfall of 725 mm. Salmon (control, fully forested) and Wights (100% cleared) are located in zone with mean annual rainfall of 1125 mm. Groundwater levels rose and the stream zone saturated area increased significantly after clearing. From analysis of this data it was evident that at a monthly time step the conceptual model framework needed to include a systematic gain/loss to storage component in order to adequately describe the observed lags between peak monthly rainfall and runoff.

Laboratory tests were undertaken to establish the formative mechanism for steps and pools in steep mountain streams. They indicated that the formation of steps and pools is associated with high intensity, low return interval events and the processes of armouring/paving and antidune formation. Lower than formative discharges give the structures their step-pool appearance, and under such discharges they are extremely stable. Step-pool streams may be modelled by a succession of artificial steps or weirs. Wooden steps were placed in a laboratory channel for this purpose, and clear water flow, clear water scour, and sediment transport tests undertaken for a range of discharges and channel slopes. Three distinct flow regimes were observed for the clear water flow and clear water scour tests. They were stable tumbling flow, unstable tumbling flow, and shooting flow. Sediment transport complicated the regimes from low transport rates. Unstable tumbling flow (clear water flow) at a low slope was shown to be caused by the breaking of standing waves at a theoretical maximum of 0.142. For higher slopes (and including clear water scour tests), unstable tumbling flow was shown to be associated with the physical system geometry preventing the submerged hydraulic jump from developing fully. However, unstable tumbling flow was also caused at lower discharges by sediment waves which were a feature of some test runs with sediment transport. Even so, unstable tumbling flow is likely to occur under field conditions only rarely. With clear water scour, the scour dimensions corresponded to the ultimate static limit. That is, no sediment remains suspended by jet action as occurs for the dynamic limit of scour. For clear water flow and clear water scour, resistance to flow may be predicted by logarithmic equations. Resistance to flow with sediment transport correlated strongly with the average scour hole size. A sudden increase in average (and maximum) velocities indicated that with sediment transport, the erosive ability of a step-pool system may increase sharply as pools become drowned by sediment. For a given discharge, increasing the sediment transport rate beyond this drowning led to net deposition, but no real increase in average velocity. With sediment transport, sediment waves and water waves occurred (independently) despite steady inputs of both water and sediment. This behaviour parallels reports of sediment movement as waves in mountain streams. This tendency toward non-uniformity of water and sediment motion suggests that such behaviour may be explicable in terms of recent advances in nonlinear thermodynamics.

Surface transient storage (STS) in stream ecosystems serve an important function in retaining nutrients and refugia for aquatic communities. Unfortunately, they can retain contaminants as well. Therefore, it is of importance to determine the residence time distribution (RTD). A RTD of a particular STS zone encompasses the time it takes for the first pulse of water to leave the STS zone, and for the mean residence time of water in that zone, among other things. The RTD of STS is also useful to subtract from the RTD of the total transient storage in streams in order to determine the hyporheic transient storage (HTS) of streams, which is difficult to measure. Currently, there is no definitive method of determining the RTD of STS. They have been determined with tracer injection alone, though this is time consuming and subject to interference from HTS. A relationship between STS physical characteristics and a RTD would be desirable, as this would characterize the time of entrainment of STS based upon a few easily measured physical parameters. This exists for groyne fields and flumes, which both have artificial STS. However, direct application of these equations to natural STS leads to errors due to simplistic geometries. The focus of this study determines RTDs in lateral STS, which is adjacent to the main channel of a stream and a significant proportion of STS, and its relationship to physically measurable parameters of lateral STS. Twenty sites throughout Oregon were each injected with NaCl to determine four residence timescales: Langmuir time (��[subscript L]), negative inverse slope of the normalized concentration curve of the primary gyre (��[subscript 1]), negative inverse slope of the normalized concentration curve of the entire STS zone (��[subscript 2]), and the mean residence time (��[subscript STS]). The RTDs of these sites were then compared to the length, width, and depth of each lateral STS zone, as well as the velocity of the adjacent main channel. This data also was used to calculate dimensionless parameters submergence, a measure of bed roughness, and k, a measure of exchange that relates ��STS to lateral STS and associated parameters. ��[subscript 1] was found to be identical to ��[subscript STS], and ��[subscript 2] could not be defined. ��[subscript STS] was found to be approximately 1.35 times ��[subscript L], the ratio of which (��[subscript L]/��[subscript STS]) is positively correlated with lateral STS submergence. ��[subscript L] and ��[subscript STS] are positively correlated with lateral STS parameters, and inversely correlated with main channel velocity. The value of k from this study was comparable to the value of k from other studies in flumes, and so there is a relationship between RTDs and lateral STS parameters. / Graduation date: 2013

Natural vegetation has been converted to land uses, such as agriculture, commercial forestry and urban use, to meet increasing human demands for food, fuel and shelter. These land use changes modify the surface conditions of an area, resulting in changes in hydrological responses. Urban land use, in particular, has a significant impact on catchment hydrology as a result of the increased impervious areas such as concrete, tar and roofs. To assess the future hydrological impacts of urban land use, the scale and location of future urban areas must be considered. The objective of this study was to assess the hydrological responses to future urban growth in the Mgeni catchment, South Africa. An urban growth model was used to generate scenarios of plausible future urban growth and these scenarios were modelled using a hydrological model to determine the hydrological responses to urban growth. The plausible future urban growth in the Mgeni catchment was modelled using the SLEUTH Urban Growth model (SLEUTH). The SLEUTH acronym stands for the input layers required for the model viz. Slope, Land use, Excluded areas, Urban Extent, Transport routes and Hillshade. SLEUTH is able to provide the scale and location of future urban growth required to assess the hydrological impacts of future urban growth. The data requirements and modelling procedure for SLEUTH is relatively simply and therefore it is well suited to a South African context. SLEUTH was calibrated and applied to the Mgeni catchment to project future urban land use. When assessing the 95-100% probability class, the results revealed that the Henley, Pietermaritzburg and Durban areas would experience the highest urban growth in the Mgeni catchment by the year 2050. The outputs of the SLEUTH Model for the Mgeni catchment showed a number of similarities to another application of SLEUTH in Cape Town. These similarities indicate the SLEUTH performs in a similar way for the two South African cities. Therefore, it was concluded that the SLEUTH Model is suitable to account for urban growth in the Mgeni catchment, as required for use in hydrological impact studies. The hydrological responses to urban growth in the Mgeni catchment were assessed using the ACRU model. The scenarios of plausible future urban growth generated by SLEUTH were overlaid with current land cover layers to generate maps of plausible future urban land use. The results showed extensive urban growth of >95% probability occurring in the Midmar, Albert Falls, Henley, Pietermaritzburg, Table Mountain, Inanda and Durban Water Management Areas (WMAs) by 2050. Increases in mean annual streamflows were observed in many of these areas; however the Henley, Pietermaritzburg and Table Mountain WMAs were shown to have greater increases in mean annual streamflow than the other areas that showed similar increases in urban growth, thus indicating that these WMAs could be particularly responsive to urban growth in the future. Furthermore, the results showed that the type of urban land use is important in determining the hydrological responses of urban land use, as the imperviousness differs between the different urban land uses. Streamflow responses were shown to be influenced by the scale and location of urban growth in the Mgeni catchment and specific areas, such as the WMAs along the Msunduzi River, were identified as potentially responsive to urban growth. Summer streamflows were indicated as being more responsive to urban land use changes than winter streamflows and increases in streamflows due to urban growth start to over-ride the impacts of other land uses which have substantial impacts on hydrological responses such as commercial forestry, and commercial sugarcane by 2050, whereas in other areas increases were mitigated by the presence of major dams. Lastly, it was shown that the type of urban land use, such as built up urban areas when compared to informal urban areas for example, have a significant impact on streamflow responses. These results are useful as they can be used to inform both water resources planning as well as urban planning to ensure that South Africa’s valuable water resources are protected. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2012.

This study involves the application of the ACRU Agrohydrological Model to a selected study catchment in the Lower Mgeni Catchment, and its discretized subcatchments, immediately downstream of the Inanda Dam. This study was initiated on the assumption that the Inanda Dam, which came into operation in 1989, would have significant impacts on the downstream (Lower Mgeni) hydrology, geomorphology and ecology. The overall aim of this study, to set up and run the ACRU model for the delimited study catchment, was successfully accomplished. This aspect of the study involved firstly, the setting up of an input database for each distributed catchment within the catchment; secondly, the processes and techniques used to translate data into hydrological information; and finally the "running" of the hydrological model, which in turn "drives" the system and simulates the catchment hydrology. Specific objectives of the study entailed the simulation of hydrology, which focussed on simulated runoff and streamflow; and sediment yield responses of the subcatchments and the total study catchment of the Lower Mgeni, with respect to gross volumes and sediment yield rates produced. The streamflow results reported indicated a season of "Iow" flow, with a monthly flowrate ranging from 1155m3s-1 to 2735m3s-1 , from April to September; and is identified and distinguished from the period of "high" flowrate, ranging from approximately 483m3s-1 to 1747m3s-1 , for the remaining months of the year. The mean annual volume for the delimited subcatchment is 22 278.5 million m3 , exceeding the annual volume required to maintain riverine and estuarine ecology, which according to DWAF (1990) is 18.5 million m3 . The simulated results of sediment yield indicate that Subcatchment 3 and 4 have the lowest sediment yield rates of 32.3 t km-2 a-1 and 32.6 t km-2 a-1 , respectively. Subcatchment 2 has the highest yield rate at the value of 617 t km-2 a-1 , while subcatchment 1 has a rate of 53.2 t km-2 a-1 . Annual sediment production in the Lower Mgeni subcatchment is 10 855.1 tons per annum with respect to gross mass, resulting in a sediment yield rate of 73.8 t km-2 a-1 . The outcomes of this study compare very favourably with other studies conducted on hydrology and sediment yield, especially those undertaken within this geographical area. It may be assumed therefore, that the results produced herein can be applied with confidence to enable appropriate planning and management of resources within this catchment. Modelling of hydrology in the Lower Mgeni is expected to contribute significantly towards meeting riverine and estuarine ecological and geomorphological streamflow requirements. It would facilitate the development of an appropriate management and dam release strategy of Inanda Dam, in order to meet these requirements. The modelling of sediment yield is expected to contribute to the development of a sustainable sandwinning policy and strategy for the Lower Mgeni, as current extraction rates exceed the annual sediment production. Once the model has been applied to a selected catchment, it has the ability to consider different scenarios, providing an invaluable tool for planning. Based on the results of this study, the ACRU model may be applied, with confidence, to other similar ungauged catchments. / Thesis (M.Sc.)-University of Natal, Durban, 2001.

Thesis (MTech (Nature Conservation))--Cape Peninsula University of Technology, 2010. / Many urban rivers are heavily engineered and no longer function naturally. The City of Cape Town has designed and implemented many stormwater and river management projects. Very little monitoring has occurred as to whether these engineering projects and remediation measures, have had a positive effect on our urban river ecosystem function. The study investigated the influence of specific engineering interventions such as the placement or rocks in stream to create weirs, gabion lining of stream channels, removal of canal walls, establishment of artificial wetlands, and approaches to urban river management, on river ecosystem function of the Keysers River, Little Lotus River, Langevlei Canal, Silvermine River, Moddergat River and Big Lotus River, as measured by specific indices including water chemistry and aquatic community structure (macroinvertebrates and diatoms).

This study investigates the hydrologic effects of urbanization in two Portland, Oregon streams through a comparison of three hydrologic metrics. Hydrologic metrics used in this study are the mean annual runoff ratio (Qa), mean seasonal runoff ratio (Qw and Qd), and the fraction of time that streamflow exceeds the mean streamflow during the year (TQmean). Additionally, the relative change in streamflow in response to storm events was examined for two watersheds. For this investigation urban development is represented by two urbanization metrics: percent impervious and road density. Descriptive and inferential statistics were used to evaluate the relationship between the hydrologic metrics and the amount of urban development in each watershed. The effect of watershed size was also investigated using nested watersheds, with watershed size ranging from 6 km2 to 138km 2. The results indicate that annual and seasonal runoff ratios have difficulty capturing the dynamic hydrologic behavior in urban watersheds. TQmean was useful at capturing the flashy behavior of the Upper Fanno watershed, however it did not perform as well in Kelley watershed possibly due to the influence of impermeable soils and steep slopes. Unexpected values for hydrologic metrics in Lower Johnson, Sycamore and Kelley watersheds could be the result water collection systems that appear to route surface water outside of their watersheds as well as permeable soils. Storm event analysis was effective at characterizing the behavior for the selected watersheds, indicating that shorter time scales may best capture the dynamic behavior of urban watersheds.

This research focuses on the estimation of the impacts of climate change on water yield, streamflow extremes, and the streamflow regimes in the Cline River Watershed, and consequently, water availability for hydropower generation in this area. The Cline River Watershed comprises the flow into Lake Abraham, the reservoir for Bighorn Dam, is part of the upper North Saskatchewan River basin (UNSRB). This objective was achieved by parameterizing the ACRU agro-hydrological modelling system. After parameterization was complete, ACRU output was calibrated and verified against available observed data, including temperature, snow water equivalent, glacier mass balance, potential evapotranspiration, and streamflow data. After ACRU was properly verified, five selected climate change scenarios to estimate impacts of climate change in this area. Overall water yields are projected to increase over time. A large shift in seasonality is likely the biggest impact climate change will have on water resources in the Cline River Watershed. / xii, 126 leaves : ill., maps ; 29 cm